Pattern detection in null model analysis

Synthesis The identification of distinctive patterns in species x site presence‐absence matrices is important for understanding meta‐community organisation. We compared the performance of a suite of null models and metrics that have been proposed to measure patterns of segregation, aggregation, nestedness, coherence, and species turnover. We found that any matrix with segregated species pairs can be re‐ordered to highlight aggregated pairs, indicating that these seemingly opposite patterns are closely related. Recently proposed classification schemes failed to correctly classify realistic matrices that included multiple co‐occurrence structures. We propose using a combination of metrics and decomposing matrix‐wide patterns into those of individual pairs of species and sites to pinpoint sources of non‐randomness. Null model analysis has been a popular tool for detecting pattern in binary presence–absence matrices, and previous tests have identified algorithms and metrics that have good statistical properties. However, the behavior of different metrics is often correlated, making it difficult to distinguish different patterns. We compared the performance of a suite of null models and metrics that have been proposed to measure patterns of segregation, aggregation, nestedness, coherence, and species turnover. We found that any matrix with segregated species pairs can be re‐ordered to highlight aggregated pairs. As a consequence, the same null model can identify a single matrix as being simultaneously aggregated, segregated or nested. These results cast doubt on previous conclusions of matrix‐wide species segregation based on the C‐score and the fixed‐fixed algorithm. Similarly, we found that recently proposed classification schemes based on patterns of coherence, nestedness, and segregation and aggregation cannot be uniquely distinguished using proposed metrics and null model algorithms. It may be necessary to use a combination of different metrics and to decompose matrix‐wide patterns into those of individual pairs of species or pairs of sites to pinpoint the sources of non‐randomness.     

人工红松林步行虫(Coleoptera:Carabidae)群落物种共存格局动态分析

揭示群落物种共存格局是群落生态学研究的重点内容之一,零模型的应用极大的促进了群落物种共存格局及其调控机制的进展,然而针对地下生物群落共存格局动态特征的研究并不多见。在帽儿山森林生态站的人工红松林内,通过5次调查取样基于零模型模拟分析小尺度空间(20 m×20 m)步行虫群落物种共存格局的动态特征。结果表明:(1)共捕获步行虫20种,2278只个体,其中Carabus billergi maoershanensis为所有调查季节数量最具优势且分布最广泛的物种,步行虫群落结构具明显的时间变异性;(2)2013年6月步行虫群落为集群性共存格局,而2014年8、10月为竞争性共存格局,基于目前的零模型指标和法则难以准确揭示其他月份是集群性还是竞争性的共存格局,但所有季节的群落均表现为明显的非随机性共存格局,这些共存格局的发现并不完全支持Diamond的群落构建机制理论;(3)所有调查季节均发现很少的显著物种对,基于更严格的检验表明群落中集群性物种对多于隔离性物种对,那些表现为显著的非随机性共存关系的物种对往往是群落内数量较大且分布广泛的优势和常见物种。表明非随机性共存格局可能是帽儿山人工红松林小尺度空间步行虫群落的常见格局,这种非随机性格局具一定的短期动态稳定性,但不同季节这种非随机性共存格局类型表现不同,群落内这些较少的显著物种对可能对群落物种共存格局具有一定的贡献。     

Pairwise null model analyses of temporal patterns of bird assemblages contradict the assumptions of competition theory

Compensatory dynamics assume inverse patterns of population dynamics of species with similar ecological resource requirements (temporal segregation). The objective of this study was to test this hypothesis on temporal samples (10–57 years) of 19 breeding bird assemblages of various habitats. We used presence/absence null model (SIM2) in combination with the C-score and Sørensen indices. The C-score index estimates the average number of checkerboards for two species, while the Sørensen index measures the qualitative similarity of co-occurrence between two species in a time series. We used pairwise null model analysis to select significant species pairs based on three selection criteria: the standard confidence interval criterion, conservative empirical Bayes mean based criterion and confidence limit based criterion. Altogether, 21 402 species pairs were analysed. The SIM2 algorithm detected from 157 to 7 segregated pairs depending on the selection criterion. The number of significant negative pairs with possible biological significance (foraging guild membership, predator–prey interactions) was far lower and represented approximately 0.0–0.3% (4–65) of pairs in a matrix. Indeed, the number of detected negative associations depended on the selection criterion. Moreover, the number of segregated pairs was negatively related to the area of the census plots and fill of the species matrix. Our results underline the minor importance of interspecific competitive interactions in temporal patterns of bird assemblages. Instead, we suggest that stochastic factors, climate or heterospecific social information may lead to more or less synchronous dynamics of bird pairs.     

Null model analyses of presence–absence matrices need a definition of independence

Tests for species interactions that involve the comparison of a statistic calculated from observed matrix of species presences and absences with the distribution of the same statistic generated from a null model have been used by ecologists for about 30 years. We argue that the validity of these tests requires a specific definition of independence. In particular, we note that an assumption that is often made is that all presence–absence matrices with the same row and column totals are equally likely if there is no interaction. However, we show using a simple model for species presences and absences without any species interactions that, in general, this assumption should be made with caution. Our model incorporates a definition of independence, allowing the computation of probabilities of different patterns in the null matrices. Other definitions of independence are possible; one of them is outlined using a new generalized linear model approach for carrying out tests applicable to different null models with or without the assumption of keeping row and column totals fixed. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Multiple regression on distance matrices: a multivariate spatial analysis tool

I explore the use of multiple regression on distance matrices (MRM), an extension of partial Mantel analysis, in spatial analysis of ecological data. MRM involves a multiple regression of a response matrix on any number of explanatory matrices, where each matrix contains distances or similarities (in terms of ecological, spatial, or other attributes) between all pair-wise combinations of n objects (sample units); tests of statistical significance are performed by permutation. The method is flexible in terms of the types of data that may be analyzed (counts, presence–absence, continuous, categorical) and the shapes of response curves. MRM offers several advantages over traditional partial Mantel analysis: (1) separating environmental distances into distinct distance matrices allows inferences to be made at the level of individual variables; (2) nonparametric or nonlinear multiple regression methods may be employed; and (3) spatial autocorrelation may be quantified and tested at different spatial scales using a series of lag matrices, each representing a geographic distance class. The MRM lag matrices model may be parameterized to yield very similar inferences regarding spatial autocorrelation as the Mantel correlogram. Unlike the correlogram, however, the lag matrices model may also include environmental distance matrices, so that spatial patterns in species abundance distances (community similarity) may be quantified while controlling for the environmental similarity between sites. Examples of spatial analyses with MRM are presented.     

A comparison of approaches for modelling the occurrence of marine animals

Approaches for modelling the distribution of animals in relation to their environment can be divided into two basic types, those which use records of absence as well as records of presence and those which use only presence records. For terrestrial species, presence–absence approaches have been found to produce models with greater predictive ability than presence-only approaches. This study compared the predictive ability of both approaches for a marine animal, the harbour porpoise (Phoceoena phocoena). Using data on the occurrence of harbour porpoises in the Sea of Hebrides, Scotland, the predictive abilities of one presence–absence approach (generalised linear modelling—GLM) and three presence-only approaches (Principal component analysis—PCA, ecological niche factor analysis—ENFA and genetic algorithm for rule-set prediction—GARP) were compared. When the predictive ability of the models was assessed using receiver operating characteristic (ROC) plots, the presence–absence approach (GLM) was found to have the greatest predictive ability. However, all approaches were found to produce models that predicted occurrence significantly better than a random model and the GLM model did not perform significantly better than ENFA and GARP. The PCA had a significantly lower predictive ability than GLM but not the other approaches. In addition, all models predicted a similar spatial distribution. Therefore, while models constructed using presence–absence approaches are likely to provide the best understanding of species distribution within a surveyed area, presence-only models can perform almost as well. However, careful consideration of the potential limitations and biases in the data, especially with regards to representativeness, is needed if the results of presence-only models are to be used for conservation and/or management purposes. Guest editor: V. D. Valavanis Essential Habitat Mapping in the Mediterranean     

Vole disturbances and plant diversity in a grassland metacommunity

We studied the disturbance associated with prairie vole burrows and its effects on grassland plant diversity at the patch (1 m²) and metacommunity (>5 ha) scales. We expected vole burrows to increase patch-scale plant species diversity by locally reducing competition for resources or creating niche opportunities that increase the presence of fugitive species. At the metacommunity scale, we expected burrows to increase resource heterogeneity and have a community composition distinct from the matrix. We measured resource variables and plant community composition in 30 paired plots representing disturbed burrows and undisturbed matrix patches in a cool-season grassland. Vole disturbance affected the mean values of nine resource variables measured and contributed more to resource heterogeneity in the metacommunity than matrix plots. Disturbance increased local plant species richness, metacommunity evenness, and the presence and abundance of fugitive species. To learn more about the contribution of burrow and matrix habitats to metacommunity diversity, we compared community similarity among burrow and matrix plots. Using Sorenson’s similarity index, which considers only presence–absence data, we found no difference in community similarity among burrows and matrix plots. Using a proportional similarity index, which considers both presence–absence and relative abundance data, we found low community similarity among burrows. Burrows appeared to shift the identity of dominant species away from the species dominant in the matrix. They also allowed subordinate species to persist in higher abundances. The patterns we observed are consistent with several diversity-maintaining mechanisms, including a successional mosaic and alternative successional trajectories. We also found evidence that prairie voles may be ecosystem engineers.     

Prediction of marine species distribution from presence–absence acoustic data: comparing the fitting efficiency and the predictive capacity of conventional and novel distribution models

The accurate representation of species distribution derived from sampled data is essential for management purposes and to underpin population modelling. Additionally, the prediction of species distribution for an expanded area, beyond the sampling area can reduce sampling costs. Here, several well-established and recently developed habitat modelling techniques are investigated in order to identify the most suitable approach to use with presence–absence acoustic data. The fitting efficiency of the modelling techniques are initially tested on the training dataset while their predictive capacity is evaluated using a verification set. For the comparison among models, Receiver Operating Characteristics (ROC), Kappa statistics, correlation and confusion matrices are used. Boosted Regression Trees (BRT) and Associative Neural Networks (ASNN), which are both within the machine learning category, outperformed the other modelling approaches tested.     

A statistical methodology for tracking long-term change in reporting rates of birds from volunteer-collected presence–absence data

The ability to track change in biodiversity is essential to guide sustainable management and meet biodiversity monitoring, evaluation and reporting requirements, yet long-term data are usually scarce. Birds Australia has developed a simple survey methodology for use by their nationwide network of volunteers; it involves the collection of data on the presence–absence of species at repeatedly visited sites. Here we present a statistical methodology for use with these binary data to examine long-term change, using as an example records from a major bioregion of eastern Australia, 1999–2007. Regression splines were employed to model trend as a smooth nonlinear function of time within a generalised linear modelling framework. Confidence intervals based on bootstrap resampling provided a basis for assessing the significance of change, and a method was incorporated for identifying important change points in the trajectory from second derivatives of the curve. The methodology proved sensitive to change and the impact of extended dry periods was evident. The populations of several woodland species were found to be in significant decline. Two composite indices to track change common to a group of birds were developed and/or adapted from the existing literature. The results confirm the usefulness of repeated 2-ha presence–absence survey data to provide insight into patterns of long-term trends in bird populations. The statistical methodology described offers a means of tracking trends and identifying important time points and is particularly useful in situations where surveys of presence–absence of species are the most efficient way to gather long-term data.     

Validation of null model tests using Neyman–Pearson hypothesis testing theory

A long-standing question in ecology is whether interspecific competition affects co-occurrence patterns of species. Null model tests of presence–absence data (NMTPAs) constitute an important approach to address the question, but different tests often give conflicting results when applied to the same data. Neyman–Pearson hypothesis testing theory provides a rigorous and well accepted framework for assessing the validity and optimality of statistical tests. Here, I treat NMTPAs within this framework, and measure the robustness and bias of 72 representative tests. My results indicate that, when restrictive assumptions are met, existing NMTPAs are adequate, but for general testing situations, the use of all existing NMTPAs is unjustified — the tests are nonrobust or biased. For many current applications of NMTPAs, restrictive assumptions appear unmet, so these results illustrate an area in which existing NMTPAs can be improved. In addition to highlighting useful improvements to existing NMTPAs, the results here provide a rigorous framework for developing improved methods.     

Body mass‐related changes in mammal community assembly patterns during the late Quaternary of North America

The late Quaternary of North America was marked by prominent ecological changes, including the end‐Pleistocene megafaunal extinction, the spread of human settlements and the rise of agriculture. Here we examine the mechanistic reasons for temporal changes in mammal species association and body size during this time period. Building upon the co‐occurrence results from Lyons et al. (2016) – wherein each species pair was classified as spatially aggregated, segregated or random – we examined body mass differences (BMD) between each species pair for each association type and time period (Late Pleistocene: 40 000 ¹⁴C–11 700 ¹⁴C ybp, Holocene: 11 700 ¹⁴C–50 ybp and Modern: 50–0 yr). In the Late Pleistocene and Holocene, the BMD of both aggregated and segregated species pairs was significantly smaller than the BMD of random pairs. These results are consistent with environmental filtering and competition as important drivers of community structure in both time periods. Modern assemblages showed a breakdown between BMD and co‐occurrence patterns: the average BMD of aggregated, segregated and random species pairs did not differ from each other. Collectively, these results indicate that the late Quaternary mammalian extinctions not only eliminated many large‐bodied species but were followed by a re‐organization of communities that altered patterns of species coexistence and associated differences in body size.     

Disentangling biotic interactions,environmental filters,and dispersal limitation as drivers of species co‐occurrence

A key focus in ecology is to search for community assembly rules. Here we compare two community modelling frameworks that integrate a combination of environmental and spatial data to identify positive and negative species associations from presence–absence matrices, and incorporate an additional comparison using joint species distribution models (JSDM). The frameworks use a dichotomous logic tree that distinguishes dispersal limitation, environmental requirements, and interspecific interactions as causes of segregated or aggregated species pairs. The first framework is based on a classical null model analysis complemented by tests of spatial arrangement and environmental characteristics of the sites occupied by the members of each species pair (Classic framework). The second framework, (SDM framework) implemented here for the first time, builds on the application of environmentally‐constrained null models (or JSDMs) to partial out the influence of the environment, and includes an analysis of the geographical configuration of species ranges to account for dispersal effects. We applied these approaches to examine plot‐level species co‐occurrence in plant communities sampled along a wide elevation gradient in the Swiss Alps. According to the frameworks, the majority of species pairs were randomly associated, and most of the non‐random positive and negative species associations could be attributed to environmental filtering and/or dispersal limitation. These patterns were partly detected also with JSDM. Biotic interactions were detected more frequently in the SDM framework, and by JSDM, than in the Classic framework. All approaches detected species aggregation more often than segregation, perhaps reflecting the important role of facilitation in stressful high‐elevation environments. Differences between the frameworks may reflect the explicit incorporation of elevational segregation in the SDM framework and the sensitivity of JSDM to the environmental data. Nevertheless, all methods have the potential to reveal general patterns of species co‐occurrence for different taxa, spatial scales, and environmental conditions.     

Habitat characteristics of crayfish (<Emphasis Type="Italic">Paranephrops planifrons</Emphasis>) in New Zealand streams using generalised additive models (GAMs)

Quantitative data on the habitat characteristics of stream crayfish have been generally lacking and competing demands on water resources has created a need to address this knowledge gap. We investigated day-time habitat relationships of stream crayfish (Paranephrops planifrons White) from 793 quadrats at 30 rivers and streams in the North Island, New Zealand to develop models of koura presence–absence and abundance. The model (stepwise GAM) included width, cover, median substrate size, edge location, velocity and depth, and correctly predicted presence–absence of crayfish (8–39 mm OCL) at 73.4% of quadrats and of young-of-the-year (YOY) ≤8 mm OCL at 83.4% of quadrats. Streams ranged from 1.6 to 11.5 m in width and the probability of finding both crayfish size classes reduced sharply as streams became wider than 6 m and as the substrate became large (i.e., boulder > 256 mm). Crayfish, particularly YOY, were most likely to be found in association with cover and at the stream edge. YOY were associated with shallow depths and fine substrates, whereas larger crayfish showed a preference for cobble substrate. Undercut banks, leaf litter, tree roots, and woody debris were strongly related to the presence–absence of crayfish. The model for crayfish abundance (log-linear Poisson GAM) explained 50% of the variation between quadrats with cover, velocity, edge location, depth, and the overall crayfish abundance at each particular stream being significant variables. Highest crayfish numbers were recorded in still or slow flowing water, with the majority occurring where velocities were below 0.4 m/s. Water depths up to 0.7 m were sampled, but highest numbers were found in depths of 0.2–0.3 m. Our presence–absence model determined variables that were significant over all streams, whereas our abundance model determined variables that were significant within streams. Use of the GAMs models enabled us to untangle the multiple factors contributing to habitat selection. Cover, velocity, and locations at the stream edge were important determinants of both presence–absence and abundance. Generally, substrate was important when comparing between streams, but not within streams, whereas depth was a significant determinant of abundance within streams, but not presence–absence between streams. Handling editor: K. Martens     

Binary matrices and checkerboard distributions of birds in the Bismarck Archipelago

Aim We examine a presence–absence matrix of the avifauna of the Bismarck Archipelago, for which the concept of competitively driven community assembly rules was formulated, to determine whether data support widespread competitive determination of geographical distributions. Location Bismarck Archipelago. Methods We obtained occurrences of 154 land and freshwater bird species on 31 islands. We calculated the observed number of checkerboards for all species pairs, for congeneric species pairs and for pairs of species within guilds, and employed randomization techniques to detect unusual co‐occurrence patterns. Results Compared with random expectations, there are excesses of checkerboard pairs within both genera and defined guilds, but a detailed examination shows that competition is a cogent possible explanation in few instances. For many checkerboard pairs, species are not widely interspersed but are regionally allopatric, which probably reflects historical biogeography and dispersal limitation. Most congeneric and intraguild checkerboards include a species classified as a supertramp; when supertramps are omitted, there are 11 congeneric checkerboards and four intraguild but heterogeneric checkerboards. Main conclusions In isolation, presence–absence matrices provide limited insight into the role of competition in structuring bird communities of the Bismarcks. A major problem is disentangling historical geography and colonization history of the archipelago from the present‐day ecology of the species. Examination of observed checkerboards from a geographically explicit perspective and with knowledge of colonization routes suggests that many checkerboards are likely to result, at least in part, from historical biogeography and supertramps. Although species may be forced into supertramp status by competition, other factors (e.g. habitat preference) may be causal, and biogeographical distributions alone cannot distinguish between causes.     

Connectivity, Probabilities and Persistence: Comparing Reserve Selection Strategies

Reserve selection methods are often based on information on species’ occurrence. This can be presence–absence data, or probabilities of occurrence estimated with species distribution models. However, the effect of the choice of distribution model on the outcome of a reserve selection method has been ignored. Here we test a range of species distribution models with three different reserve selection methods. The distribution models had different combinations of variables related to habitat quality and connectivity (which incorporates the effect of spatial habitat configuration on species occurrence). The reserve selection methods included (i) a minimum set approach without spatial considerations; (ii) a clustering reserve selection method; and (iii) a dynamic approach where probabilities of occurrence are re-evaluated according to the spatial pattern of selected sites. The sets of selected reserves were assessed by re-computing species probability of occurrence in reserves using the best probability model and assuming loss of non-selected habitat. The results show that particular choices of distribution model and selection method may lead to reserves that overestimate the achieved target; in other words, species may seem to be represented but the reserve network may actually not be able to support them in the long-term. Instead, the use of models that incorporated connectivity as a variable resulted in the selection of aggregated reserves with higher potential for species long-term persistence. As reserve design aims at the long-term protection of species, it is important to be aware of the uncertainties related to model and method choice and their implications.     

Floodplain lake fish assemblages in the Amazon River: directions in conservation biology

The Neotropical region is renowned for its high biodiversity, and the Amazon River basin contains the highest number of fish species of any river system in the world. In recent years, habitat fragmentation and exploitation of biotic resources have threatened biological integrity and provoked to need for sustainable management and conservation of the Amazon River system. We studied 36 floodplain lakes along 2000 km of the Amazon River. The fish assemblages associated with flood forests are moderately diverse, with low species dominance and reduced populations. To detect nestedness of fish assemblage composition in floodplain lakes, a nested subset analysis was performed on species presence–absence. The incidence matrix (species × lakes) was maximally packed using the Nestedness Temperature Calculator software. The results of ranking lakes and species allow us to establish targets for conservation. Such strategy for sustainable management should be focused on maintaining the Amazonian biodiversity.     

Relating macroinvertebrate community structure to environmental characteristics and sediment contamination at the scale of the St. Lawrence River

There is still no assessment of the impact of sediment chemicals and environmental conditions on macroinvertebrates at the scale of the St. Lawrence River. In order to assess these impacts in the fluvial section of the St. Lawrence River including the Montreal harbour, the community structure of macroinvertebrates using different taxonomic aggregations (genus and family) and taxa attributes (abundance, presence–absence, indicator taxa) was assessed. The goal of the study was to determine the indicator taxa of macroinvertebrates along the fluvial continuum and relate changes in macroinvertebrate community to sediment chemical conditions and environmental characteristics of habitats using variance partitioning. This study also evaluated which taxonomic level and taxa attributes of macroinvertebrates were the most suitable for bioassessment of quality of sediments and habitat environment in the St. Lawrence River. Four different macroinvertebrate assemblages were found distributed along the fluvial continuum using either abundance or presence–absence data and genus or family levels. Indicator taxa characteristic of the different macroinvertebrate communities were associated with the sediment contamination gradient. However, habitat environmental characteristics (water masses, sulphur and DOC in sediments) had more influence on macroinvertebrate assemblages than sediment contamination. Our study confirms that family level analysis can give information comparable to the genus level analysis using presence–absence or abundance of macroinvertebrates, yet a higher number of indicator taxa were detected at the genus level.     

Breeding behavior and mating system in relation to body size in <Emphasis Type="Italic">Rana chensinensis</Emphasis>, a temperate frog endemic to northern China

Knowledge of how mating success is related to body size may provide insight into the evolution of social systems. This study investigated the mating system and relevant social behavior of a temperate anuran (Rana chensinensis) at three localities in northern China. During chorusing, males aggregated and persisted in the communal spawning ponds with a density of 10–26 frogs per m² water area and operational sex ratio of 15–28 males to 1 female. The males frequently grabbed any conspecifics they encountered, releasing the grip if the individual was a male, but holding the grip when it was a female. A significant positive relationship between male and female body lengths of pairs in amplexus was detected from all the sites during the five breeding seasons, but the average correlation coefficients of determination of 18% indicated that the size-assortative mating constituted a minority of the species’ mating system. Pairing probability decreased with increased size differences between sexes, which could have prevented a few adult animals in a population from forming pairs. The strong aggregation of males could limit the opportunities for large males to exhibit contest advantages in mating and for both sexes to choose a large mate. Infrequent occurrence of scramble competition (averaging 8% of the recorded amplectant pairs) and the lack of observed takeovers, probably because of reclusion of amplectant pairs at the bottom of spawning ponds, suggested a weak role of amplexus displacement in generating non-random mating. Our results suggest that although non-random mating with respect to body size may be expected in explosive breeders, the social environment characterized by male aggregation may also impose weak selective pressure for the evolution of non-random mating.     

Pairwise versus presence–absence approaches for analysing biodiversity patterns

It has been proposed that the study of co‐occurrence of species, which is traditionally performed using full presence–absence matrices of sets of many species, could benefit from simply testing for random co‐occurrence between pairs of species, and that use of a full presence–absence matrix is tantamount to regarding it as having some real ecological identity. Here I argue that although there are valid questions that can be answered using a pairwise approach, there are many others that naturally require the analysis of entire sets of species in a joint way, as provided for through the use of full presence–absence matrices. Moreover, there are theoretical and mathematical advantages to the use of presence–absence matrices, a few of which are briefly discussed in this short note.     

Co‐occurrence patterns in a diverse arboreal ant community are explained more by competition than habitat requirements

A major goal of community ecology is to identify the patterns of species associations and the processes that shape them. Arboreal ants are extremely diverse and abundant, making them an interesting and valuable group for tackling this issue. Numerous studies have used observational data of species co‐occurrence patterns to infer underlying assembly processes, but the complexity of these communities has resulted in few solid conclusions. This study takes advantage of an observational dataset that is unusually well‐structured with respect to habitat attributes (tree species, tree sizes, and vegetation structure), to disentangle different factors influencing community organization. In particular, this study assesses the potential role of interspecific competition and habitat selection on the distribution patterns of an arboreal ant community by incorporating habitat attributes into the co‐occurrence analyses. These findings are then contrasted against species traits, to explore functional explanations for the identified community patterns. We ran a suite of null models, first accounting only for the species incidence in the community and later incorporating habitat attributes in the null models. We performed analyses with all the species in the community and then with only the most common species using both a matrix‐level approach and a pairwise‐level approach. The co‐occurrence patterns did not differ from randomness in the matrix‐level approach accounting for all ant species in the community. However, a segregated pattern was detected for the most common ant species. Moreover, with the pairwise approach, we found a significant number of negative and positive pairs of species associations. Most of the segregated associations appear to be explained by competitive interactions between species, not habitat affiliations. This was supported by comparisons of species traits for significantly associated pairs. These results suggest that competition is the most important influence on the distribution patterns of arboreal ants within the focal community. Habitat attributes, in contrast, showed no significant influence on the matrix‐wide results and affected only a few associations. In addition, the segregated pairs shared more biological characteristic in common than the aggregated and random ones.